AA-trans: Core attention aggregating transformer with information entropy selector for fine-grained visual classification.
Qi WangJianJun WangHongyu DengXue WuYazhou WangGefei HaoPublished in: Pattern Recognit. (2023)
Keyphrases
- fine grained
- information entropy
- visual classification
- coarse grained
- discernibility matrix
- visual recognition
- rough sets
- image classification
- rough set theory
- access control
- decision table
- multiple features
- scene classification
- bag of features
- fuzzy logic
- attribute values
- linear classifiers
- visual attention
- combining multiple
- neural network
- face recognition
- object recognition
- high level
- computer vision